Quantitative real time polymerase chain reaction (qPCR) has become a widely used tool to examine gene expression levels. Reliable quantification, however, depends on a proper normalization strategy. Normalization with multiple reference genes is becoming the standard, although the most suitable reference genes depend on the applied treatment as well as the tissue or cell type studied. In this study the stability of various reference genes was investigated in cultures of oligodendrocytes derived from either mature or neonatal rats, the latter also in the presence of the liver X receptor (LXR) agonist.The expression stability of ten commonly used reference genes (HPRT, GAPDH, 18S, ActB, CycA, Tbp, Rpl13A, YWHAZ, HMBS, Pgk1) was analyzed using geNorm and NormFinder.When comparing the different types of cell cultures, Rpl13A, CycA, Pgk1 and YWHAZ were identified as most stable genes. After LXR agonist treatment, CycA, Pgk1 and Rpl13A were found to be the most stable by both geNorm and NormFinder. HMBS and the commonly used housekeeping genes GAPDH and 18S turned out to be the most variable according to geNorm and NormFinder. In conclusion, the use of multiple reference genes, instead of only one, in qPCR experiments with rat oligodendrocytes is strongly advised and standard housekeeping genes such as GAPDH and 18S are not recommended as they appear to be relatively unstable under the experimental conditions used. Reference gene selection should always be performed for each individual experiment, since useful reference genes are very specific for every situation.